Research on Navigation Path Extraction and Obstacle Avoidance Strategy for Pusher Robot in Dairy Farm

Existing push robots mainly use magnetic induction technology. These devices are susceptible to external electromagnetic interference and have a low degree of intelligence. To make up for the insufficiency existing material pushing robots, at same time solve problems labor-intensive, inability in night, etc., this study, an autonomous navigation pusher robot based on 3D lidar is designed, obstacle avoidance strategy improved artificial potential field method proposed. Firstly, point cloud data barn collected by self-designed robot, area interest extracted using direct-pass filtering algorithm, segmented height threshold. Secondly, Least-Squares Method (LSM) Random Sample Consensus (RANSAC) were used extract fence lines, then boundary contour features projection onto ground. Finally, target influence factor added repulsive function determine principle optimal selection parameters direction, clarify robot. It can verify effect algorithm. The experimental results showed that under three different environments: no noise, Gaussian lines RANSAC. Taking change slope as indicator, obtained about ?0.058, 0.058, ?0.061, respectively. RANSAC has less variation compared no-noise group. Compared with LSM, extraction did not significantly, indicating certain resistance various noises, but performs better real-time performance. simulation actual test show select reasonable force directions. optimized path increases shortest distance from 0.18 0.41 m, where average 0.059 s, standard deviation 0.007 s. This shows optimization optimize real avoid obstacles, basically meet requirements security performance, effectively local minimum problem. research will provide corresponding technical references overcome process operation complex open scenarios.